The impact of naturalistically produced mental state words on infants’ language development

• In Event: Poster Session 08
  In Poster Session: PS 08 Topic - Language, Communication

Abstract

Research has established the importance of mental state words (MSW) on young children’s cognitive, socioemotional, and language development (Farkas et al., 2018; Ensor & Hughes, 2008). Studies based on observational free-play and elicitation tasks have shown that MSW in speech input play an important role in infants’ overall vocabulary development at a later age (Farkas et al., 2018; Laranjo & Bernier, 2013). Specific types of MSW such as volition (want), perception (see), cognition (think), and disposition (happy) also appear to support infants’ language development differently over time (Olson & Masur, 2020). However, it is unclear if parents of young infants naturally produce MSW in their home environments and whether naturally produced MSW exhibit the same effects on infants’ language development. The present study recorded families across two days using the LENA device and tested how the presence of MSW in everyday interactions may impact infants’ language development.

Forty-eight infants (25 males) and their families participated in our study at 12-, 15-, and 18-months. At 12-months, we obtained measures of children’s input using LENA recordings from the home. All households were monolingual native-English speakers. Infants’ vocabulary knowledge was also collected at 12-, 15-, and 18-months using the MCDI. At 18-months, infants completed a lab-based novel word learning task. For each infant, audio recordings were transcribed in CLAN. Counts of MSW tokens on the transcripts were identified and subcategorized. 570 full sentences containing MSW were coded across infants. Maternal education (MatEd) was collected to control for SES-related effects based on previous research5.

We first examined the relationships of MatEd and MSW by conducting a series of mixed effects models. Overall MSW use and the amount of cognition, perception, and volition words were significantly predicted by MatEd (all ps <.001). We then evaluated the relationships of MSW use and MatEd on infants’ language development by fitting a series of multilevel models. After controlling for MatEd, the amount of total MSW explained additional unique variance in infants’ receptive vocabulary at 12-months, productive vocabulary at 15- and 18-months, and novel word learning accuracy at 18-months (all ps<.001). The specific types of MSW did not explain additional variance in infants’ vocabulary and novel word learning above and beyond MatEd (all ps>.05). LENA produced estimates on the overall recording did not correlate with MSW counts.

We found that even though all MSW and its specific types were significantly related to MatEd, only caregivers’ overall MSW use at 12 months predicted infants’ concurrent and future language development above and beyond maternal education. This finding contrasts with previous work that has found different types of MSW differently support infants’ vocabulary development in the second year of life (Olson & Masur, 2020). We also show that the total number of MSW supports infants’ novel word learning 6 months later. Together, these patterns suggest that naturalistically produced speech may exhibit characteristics distinctive from those captured in a controlled setting. Implications for how MSW use may prime and facilitate infants’ vocabulary development will also be discussed.

Group Authors

Tianlin Wang, Rong Huang, Wenqian Robertson, Elie ChingYen Yu

Authors

• Tianlin Wang, State University of New York at Albany (SUNY Albany)
  Presenting Author

• Rong Huang, University of Connecticut
  Non-Presenting Author

• Elie ChingYen Yu, State University of New York at Albany (SUNY Albany)
  Non-Presenting Author

• Wenqian Robertson, State University of New York at Albany
  Non-Presenting Author